marcelo aguiar and swee hong chan cornell universitysweehong/pdfs/presentation_gscc.pdf · marcelo...

Post on 15-Oct-2020

13 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

Toric arrangements that come from graphsMarcelo Aguiar and Swee Hong Chan

Cornell University

Toric arrangements

Layman’s terms: lines on a donut.

Studied in connection to Kostant partition functions (DeConcini-Procesi ‘05), arithmetic matroids (Moci ‘12),arithmetic Tutte polynomial (D’Adderio-Moci ‘13), etc.

This talk is about toric arrangements that are built fromgraphs.

Motivation

The current study of toric graphic arrangements is mainlyfocused on the case of the standard torus.

We study graphic arrangements on two other types of tori,the coweight torus and the coroot torus.

We will see that these two arrangements tell us new thingsabout the acyclic orientations of the input graph.

ToriLet V be a real vector space.

A lattice L is the integer-span of a basis of V .The associated torus is the quotient T := V /L.

◦◦

◦◦

◦ ◦◦◦◦

◦◦

◦◦

◦ ◦

◦ ◦

◦◦

◦◦

◦ ◦◦◦◦

◦◦

◦◦

◦ ◦

◦ ◦L

T

Graphic arrangementsLet G be a simple connected graph.A(G ) is called the linear graphic arrangement.A(G ) is called the affine graphic arrangement.

1

2 3

e2−e3

e1−e3

e1−e2

K3

• •1√2

A(K3) A(K3)

Toric graphic arrangements, example 1

◦◦

◦◦

◦ ◦◦◦◦

◦◦

◦◦

◦ ◦

◦ ◦

◦◦

◦◦

◦ ◦◦◦◦

◦◦

◦◦

◦ ◦

◦ ◦

L1 A(K3)

A(K3, L1)

Toric graphic arrangements, example 2

◦◦

◦◦

◦ ◦◦◦◦

◦◦

◦◦

◦ ◦

◦ ◦

••

••

••

◦◦

◦◦

◦•

◦◦

◦◦

◦•

L1

A(K3)

A(K3, L2)

Toric graphic arrangements, example 2

◦◦

◦◦

◦ ◦◦◦◦

◦◦

◦◦

◦ ◦

◦ ◦

••

••

◦◦

◦◦

◦•

◦◦

◦◦

◦•

A(K3)

A(K3, L2)

Toric graphic arrangements, example 2

◦◦

◦◦

◦ ◦◦◦◦

◦◦

◦◦

◦ ◦

◦ ◦

••

••

◦◦

◦◦

◦•

◦◦

◦◦

◦•

L2 A(K3)

A(K3, L2)

Root system of type A

Vn := {x ∈ Rn | x1 + · · ·+ xn = 0} (ambient space);

An−1 := {ei − ej | 1 ≤ i < j ≤ n} (root system of type A);

ZAn−1 := Z{1/n(e1 + · · ·+ en)− ei | 1 ≤ i ≤ n} ((co)weight lattice);

ZAn−1 := Z{ei − ej | 1 ≤ i < j ≤ n} ((co)root lattice).

e2−e3

e1−e3

e1−e2

e3−e2

e3−e1

e2−e1• •

◦◦

◦◦

◦ ◦◦◦◦

◦◦

◦◦

◦ ◦

◦ ◦

◦◦

◦◦

◦◦

◦◦◦

◦◦◦◦◦

◦◦◦◦◦

V3 ZA2 ZA2

One graph, three toric arrangements

The standard arrangement: V = Rn, L = Zn.

The coweight arrangement: V = Vn, L = ZAn−1.

The coroot arrangement: V = Vn, L = ZAn−1.

A(K3, ZA2) A(K3,ZA2)

Toric chambers

A(K3, ZA2)

A(K3,ZA2)

Acyclic orientationsRecall the bijection of Greene and Zaslavsky (’83):

Chambers of A(G )↔ Acyclic orientations of G

1

2 3

1

2 31

2 3

1

2 3

1

2 3

1

2 3

A(K3)

Coweight Voronoi cells

◦ ◦

◦◦

◦◦

◦ ◦

◦ ◦

1

2 3

1

2 31

2 3

1

2 3

1

2 3

1

2 3

ZA2 A(K3)

Coweight Voronoi relation

1

2 3

1

2 31

2 3

1

2 3

1

2 3

1

2 3

ZA2

Two orientations that are projected to the same toric chamber areVoronoi equivalent.

Coweight Voronoi relation

1

2 3

1

2 31

2 3

1

2 3

1

2 3

1

2 3

•ZA2

Two orientations that are projected to the same toric chamber areVoronoi equivalent.

Coroot Voronoi cells

◦◦

◦◦

◦◦

◦◦◦

◦◦

◦◦

◦◦

◦◦

1

2 3

1

2 31

2 3

1

2 3

1

2 3

1

2 3

ZA2 A(K3)

Coroot Voronoi relation

1

2 3

1

2 31

2 3

1

2 3

1

2 3

1

2 3

ZA2

No two distinct acyclic orientations are Voronoi equivalent.

Coroot Voronoi relation

1

2 3

1

2 31

2 3

1

2 3

1

2 3

1

2 3

•ZA2

No two distinct acyclic orientations are Voronoi equivalent.

Combinatorial description for coweight Voronoi equivalence

The relation is source-to-sink flip.

→•

Studied by

Mosesjan (‘72) and Pretzel (‘86) in combinatorics;Eriksson and Eriksson (‘09), and Speyer (‘09) in connection toconjugacy of Coxeter elements;Develin, Macauley and Reiner (‘16) in the context of toricarrangements.

It also arises in connection to sandpile groups and chip-firing.

Combinatorial description for coroot Voronoi equivalence

The relation has several equivalent descriptions.

One is source-sink exchange.

↔•

Another one is n-step source-to-sink flip.

◦◦

◦→

◦◦

◦→• ◦

◦ ◦→•

◦◦

◦→◦

◦◦

THANK YOU!

top related